Manufacturology

Creating pressure for process innovation with production systems models

This post is a slight rewrite of the abstract for a pre­sen­ta­tion that I’m prepar­ing for a mini–conference next week.

Most niche mar­ket steel pro­duc­ers, like for exam­ple all major Swedish steel com­pa­nies, would likely be able to make sub­stan­tial pro­duc­tiv­ity improve­ments with pro­cess­ing tech­nol­ogy that per­mit cost effi­cient low vol­ume pro­duc­tion in the pres­ence of high prod­uct vari­ety. Cur­rent pro­cess­ing tech­nol­ogy is not opti­mal since it has been designed for high pro­duc­tiv­ity in the case of high-volume pro­duc­tion and low prod­uct variety.

I make these claims based on the results of my own analy­ses. Over the last few years I’ve devel­oped a num­ber of steel pro­duc­tion sys­tems mod­els in order to study how the steel­mak­ing, con­tin­u­ous cast­ing and hot rolling processes inter­act dynam­i­cally dur­ing oper­a­tion. The result­ing behav­iour depends on, among other things, the com­bi­na­tion of cus­tomer order pat­terns, pro­cess­ing tech­nol­ogy and pro­duc­tion con­trol strategies.

Why is pro­duc­tion sys­tems mod­el­ling and sim­u­la­tion a good idea? I think the most impor­tant rea­son is that pro­duc­tion sys­tems mod­els aid, and cre­ate pres­sure for, process innovation.

My expe­ri­ence is that cur­rent steel research is mainly focused on fun­da­men­tal mate­ri­als sci­ence and improve­ment of exist­ing pro­cess­ing tech­nol­ogy. Addi­tion­ally, the pre­vail­ing assump­tion is that pro­duc­tion in large batches is most effi­cient. The result is pro­cess­ing tech­nol­ogy that is designed for high-volume low prod­uct vari­ety production.

Research on what type of capa­bil­i­ties that yield opti­mal pro­duc­tiv­ity under actual mar­ket­ing con­di­tions, as well as research on how such opti­mal capa­bil­i­ties can be realised, is vir­tu­ally nonexistent.

Dynamic inter­ac­tion between pro­cess­ing steps result e.g. from time lags, infor­ma­tion feed­back and accu­mu­la­tion of in-process inven­tory. Design­ing mod­els that account for dynamic aspect is impor­tant because much “fric­tion” appears in the inter­faces between process steps.

• Process cost mod­els esti­mate pro­cess­ing cost based on resource con­sump­tion dur­ing operation.
• Dynamic process cost mod­els account for dynamic inter­ac­tion between pro­cess­ing steps.

How­ever, mod­els do not need to be very com­plex. They are typ­i­cally com­posed of linked process mod­els that are sim­pler than most mod­els used in mate­ri­als sci­ence, but more com­plex than mod­els used in eco­nom­ics and deci­sion sci­ences. The pre­ci­sion of mod­els may be unsat­is­fac­tory for a mate­ri­als sci­en­tist, but pro­vide much bet­ter esti­mates than tra­di­tional eco­nomic cost models.

I believe that more wide­spread use of cross dis­ci­pli­nary pro­duc­tion sys­tems cost mod­el­ling within the field of mate­ri­als sci­ence and process engi­neer­ing would cre­ate pres­sure towards research and devel­op­ment of pro­cess­ing tech­nolo­gies capa­ble of cost effi­cient small-batch pro­duc­tion. Such flex­i­ble pro­cess­ing tech­nolo­gies could simul­ta­ne­ously yield improved pro­duc­tiv­ity and increased envi­ron­men­tal sustainability.

Be Socia­ble, Share!

•